Open-Source Code & Hardware

Apollo4 firmware for the A3EM sensor board — the base configuration layer, recording modes, and low-power peripheral management described on the Research page.

Firmware integration of on-device machine learning — quantized classifiers and adaptive filtering running directly on the sensor board.

AI training and deployment tools for the A3EM project — model training, quantization, and export pipelines feeding the embedded classifiers.

Research notebooks for the VAE-based encoder and online clustering novelty-detection approach used in adaptive data collection.

Python-based configuration dashboard for A3EM deployments — the graphical provisioning tool referenced in the firmware's three-tier configuration hierarchy.

Web-based configuration dashboard for A3EM deployments — a browser-based alternative for provisioning and managing devices.

A tracked fork of the BirdNET Analyzer, underpinning the project's code-free custom-classifier workflow built around its GUI and Raven Pro metadata conventions.